Model comparison

MiniMax-M2.7 vs Qwen2.5 72B Instruct

MiniMax-M2.7 is the stronger model overall, scoring 37.7 to 31.9 on the Noometry Index.

Last verified October 10, 2026. 19 shared benchmarks.

MiniMax-M2.7

MiniMax

37.7

Rank #196 Confirmed

Qwen2.5 72B Instruct

Alibaba (Qwen)

31.9

Rank #267 Confirmed

Summary

They share 19 benchmarks with published results for both. MiniMax-M2.7 scores higher in 8 categories and Qwen2.5 72B Instruct in 1 category; 9 gaps are clear of the uncertainty.
The widest gap is in writing & preference, where MiniMax-M2.7 leads 58.9 to 46.7.
The biggest single-benchmark swing is WeirdML: 37% for MiniMax-M2.7 and 16% for Qwen2.5 72B Instruct.
MiniMax-M2.7 is cheaper at $0.30 / $1.20 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
MiniMax-M2.7 accepts more context: 205K tokens versus 131K.

Side by side

MiniMax-M2.7 and Qwen2.5 72B Instruct specifications
	MiniMax-M2.7	Qwen2.5 72B Instruct
Provider	MiniMax	Alibaba (Qwen)
Noometry Index	37.7	31.9
Released	2026-03-18	2024-09
Weights	Open	Open
Context window	205K	131K
Max output	131K	8K
Input $ / M tokens	$0.30	$1.40
Output $ / M tokens	$1.20	$5.60
Results tracked	30	43

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding MiniMax-M2.7 leads

MiniMax-M2.7: 41.8 (#120), Qwen2.5 72B Instruct: 33.2 (#260)

Coding benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
WeirdML	37%	16%
LMArena Coding	1454	1292
LMArena WebDev	1398	—
SciCode	47%	—
BigCodeBench Instruct	—	45.8%
BigCodeBench Complete	—	55.9%
ALE-Bench	599.25	—

Agentic & Tool Use MiniMax-M2.7 leads

MiniMax-M2.7: 25.1 (#111), Qwen2.5 72B Instruct: 22.1 (#133)

Agentic & Tool Use benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
Terminal-Bench	45.1%	—
TheAgentCompany	—	5.7%
BALROG	—	16.2%
ExploitBench	13.3%	—
GBAEval	0%	—
METR Time Horizons	—	35.8%

Reasoning Qwen2.5 72B Instruct leads

MiniMax-M2.7: 19.7 (#253), Qwen2.5 72B Instruct: 22.3 (#199)

Reasoning benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Hard Prompts	1422	1271
Epoch Capabilities Index	145.85	129
NYT Connections (extended)	24.7%	—
CritPt	0.6%	—
Thematic Generalization	39.3%	—
DTBench	—	62.9%
LMCA	—	13.4%
BIG-Bench Hard	—	79.8%
ForecastBench	—	57.5
HellaSwag	—	84.8%
PIQA	—	82.6%
WinoGrande	—	82.3%

Math MiniMax-M2.7 leads

MiniMax-M2.7: 25.9 (#263), Qwen2.5 72B Instruct: 19.3 (#287)

Math benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Math	1420	1283
OTIS Mock AIME 2024-2025	—	8.1%
ProofBench	3%	—
Omni-MATH	—	33%
MATH Level 5	—	63.2%

Knowledge MiniMax-M2.7 leads

MiniMax-M2.7: 37.7 (#152), Qwen2.5 72B Instruct: 27.0 (#253)

Knowledge benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Expert	1444	1245
GPQA Diamond	—	49.1%
MMLU-Pro	—	63.1%
Confabulations	—	19.1%
Vectara Hallucination Rate	12.9%	—
GPQA (HELM)	—	42.6%
ARC (AI2) Challenge	—	94.5%
MMLU	—	85.3%
TriviaQA	—	71.9%

Multilingual MiniMax-M2.7 leads

MiniMax-M2.7: 50.3 (#123), Qwen2.5 72B Instruct: 41.0 (#213)

Multilingual benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Non-English	1382	1252
LMArena Chinese	1441	1272
LMArena French	1421	1280
LMArena German	1398	1234
LMArena Japanese	1262	1180
LMArena Korean	1313	1188
LMArena Russian	1383	1264
LMArena Spanish	1403	1256

Instruction Following MiniMax-M2.7 leads

MiniMax-M2.7: 74.1 (#103), Qwen2.5 72B Instruct: 65.5 (#221)

Instruction Following benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Instruction Following	1405	1254
IFEval	—	80.6%

Long Context MiniMax-M2.7 leads

MiniMax-M2.7: 43.3 (#99), Qwen2.5 72B Instruct: 38.9 (#188)

Long Context benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Longer Query	1419	1282

Writing & Preference MiniMax-M2.7 leads

MiniMax-M2.7: 58.9 (#112), Qwen2.5 72B Instruct: 46.7 (#215)

Writing & Preference benchmarks
Benchmark	MiniMax-M2.7	Qwen2.5 72B Instruct
LMArena Text	1405	1269
LMArena Creative Writing	1354	1221
LMArena Multi-Turn	1412	1272
WildBench	—	80.2%

Frequently asked questions

Is MiniMax-M2.7 better than Qwen2.5 72B Instruct?

MiniMax-M2.7 is the stronger model overall, scoring 37.7 to 31.9 on the Noometry Index.

Which is cheaper, MiniMax-M2.7 or Qwen2.5 72B Instruct?

MiniMax-M2.7 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.

Is MiniMax-M2.7 or Qwen2.5 72B Instruct better for coding?

MiniMax-M2.7 scores higher on coding benchmarks: 41.8 versus 33.2 in the Noometry coding category.

Which has the bigger context window?

MiniMax-M2.7 does, with 205K tokens against 131K.

How many benchmarks do MiniMax-M2.7 and Qwen2.5 72B Instruct share?

19 benchmarks have published results for both models. MiniMax-M2.7 has 30 scored results on Noometry and Qwen2.5 72B Instruct has 43.